Platform neutrality and the question of media effects

4.104 The ALRC is proposing that policies and regulations applying under the National Classification Scheme should reflect, to the maximum degree possible, the principle of platform neutrality. Classification should focus on media content rather than platforms or delivery technologies. In the context of media convergence, we have argued that attempts to apply different regulatory frameworks to media based upon their delivery platform has proven to be unsustainable over time, and has generated significant distortions in classification outcomes.

4.105 A related issue is the possible effects that different forms of media may have on the behaviour of individuals. The lengthy debate about whether to introduce an R 18+ classification for computer games, and the distortions and anomalies that emerged in the Australian games market arising from the absence of such a classification, has drawn attention to the problems that can arise from assumptions about media effects.

4.106 The literature on whether particular media content has effects on those who consume it is voluminous. The relationship between media violence and violence in society is perhaps the most researched topic in media and communications, with studies dating back as far as the 1930s. Research into the relationship between television and violence has been particularly prominent since the mid-1950s, after the United States (US) Congressional hearings of 1952 and 1955.^[83]

4.107 Research has often been triggered by particular events, such as riots and political assassinations in the US in the 1960s, the Columbine school shootings in the US in 1989, or—in the Australian context—the aftermath of the killing of 35 people at Port Arthur, Tasmania, by Martin Bryant in 1996. More recently, both the Oslo shootings and the London riots in 2011 acted as prompts for debate about the influence of violent video games and social media respectively.

4.108 An overview of debates as they relate to the influence of media on behaviour can be found in a 2008 special issue of American Behavioural Scientist. Those who argue that the effects of sustained exposure to violent media on children are significant, generally point to three classes of effects:

• Aggression: Viewing televised violence can lead to increases in aggressive behaviour and/or changes in attitudes and values favo[u]ring the use of aggression to solve conflicts.

• Desensitization: Extensive violence viewing may lead to decreased sensitivity to violence and a greater willingness to tolerate increasing levels of violence in society.

• Fear: Extensive exposure to television violence may produce the mean world syndrome effect, in which viewers overestimate their risk of victimisation.^[84]

4.109 Similar observations have been made by the Australian Psychological Society Ltd, which observed that ‘[e]xposure to violent television can and does influence children’s feelings, attitudes and behaviour’, and that ‘prolonged exposure to television violence is one of a number of factors which lead to children being more likely to display aggressive behaviour in both the short-term and the long-term’.^[85] Among those submissions who commented on this issue, Family Voice Australia referred to studies concerning violent video games and their impact on children, and the Australian Council for Children and the Media also provided references to relevant studies.^[86]

4.110 There has also been considerable questioning of claims about strong media effects on individual behaviour. In an overview of 50 years of research on media violence, Professor Barrie Gunter points to six factors that qualify strong claims being made about the impact of media violence that draw upon empirical research.^[87]

4.111 First, whether the studies took place in an experimental setting or were based upon ‘real world’ data. It has been noted that three-quarters of studies undertaken have been by psychologists, and about half of these have been laboratory-type experiments.^[88] These are open to criticism that they do not replicate ‘real world’ media consumption practices, and that participants go into such experiments with a pre-conceived idea of what researchers are expecting to find.

4.112 Secondly, the use of experimental methods that seek to uncover cause-effect relationships can neglect the degree to which, if media violence does impact upon behaviour, the relationship is more likely to be longer-term and cumulative rather than short-term and immediate. There is considerably less longitudinal data available on these questions as compared to experimental studies, and the meta-analytic studies (those that draw together the findings of multiple studies) find only weak correlations at best.

4.113 Thirdly, in so far as there has been a link established, it has generally been associated with those of lower socio-economic status backgrounds, or particular racial minorities. As researchers such as George Comstock observe, such groups also ‘consist of individuals who already face considerable challenges in coping with everyday life’ including a greater likelihood of conflict with authority and the law.^[89] Given that the relationships are multi-causal, this leaves open the question as to whether the media-centric focus of effects research occurs at the expense of considering other relevant socio-cultural and socio-economic factors.

4.114 Fourthly, the research literature is dominated by studies looking at the potentially harmful effects of various forms of media exposure, with few studies considering neutral or even positive consequences of exposure. For instance, if media consumers are clear about the difference between media violence and real violence, then the portrayal of violence can be an entirely legitimate form of storytelling—and one with a very long history—particularly if it also conveys a message that aggressive or anti-social behaviour can have negative consequences for its perpetrators.

4.115 Fifthly, the question of whether media consumers in general, and children in particular, differentiate between media violence and real violence can be neglected in experimental studies. Professor Stuart Cunningham has made the point, in relation to work undertaken by the Australian Broadcasting Tribunal on media violence in the early 1990s, that those surveyed were more likely to be disturbed by violent scenes witnessed on television news broadcasts than by fictionalised portrayals of violence in feature films or television dramas.^[90]

4.116 The sixth and final point: the risk of assuming that the link between media violence and social violence has been proven is that ‘an oversimplified position … can lead to political misrepresentation of media effects, with unreasonable requests for tighter controls over media content, scheduling, and transmissions’.^[91]

4.117 The argument presented here is not that there are no effects of media on individual behaviour. Gunter concludes that ‘certain forms of media violence can exert certain kinds of effects on some consumers some of the time’,^[92] and Dr. Andy Ruddock from Monash University has identified particular contexts where particular media consumers actively use media to achieve certain kinds of effects.^[93] It is, rather, to note that there are many and varied results from these studies, and that this evidence base has not generated clearer findings over time.

4.118 This would suggest that there are inherent difficulties in making recommendations about content classification policy and regulation based on claims of media effects on human behaviour. This conclusion is similar to that reached by the Australian Government Attorney-General’s Department in its literature review on the impact of playing violent video games (VVGs) on aggression:

Significant harmful effects from VVGs have not been persuasively proven or disproven. There is some consensus that VVGs may be harmful to certain populations, such as people with aggressive and psychotic personality traits. Overall, most studies have consistently shown a small statistical effect of VVG exposure on aggressive behaviour, but there are problems with these findings that reduce their policy relevance. Overall … research into the effects of VVGs on aggression is contested and inconclusive.^[94]

4.119 The ALRC is of the view that the evidence on media effects on individual behaviour is sufficiently ambiguous that it would advise against applying different classification criteria or restrictions to different platforms on this basis. As a result, a content-based approach to classification is the approach adopted in this Report.